Subject: Re: OT: CAPE's Predictive Power
If you take a return of precisely 10 years, then the result you get is at the mercy of the (more or less random) valuation level on the precise ending anniversary.
Yes, I averaged in another section, it's a good idea. But for 1995 onwards, the results are great whether you average or not.

But I find results in previous 15 year periods (1995 to now is 18 years, I used 15 as a nice number for a sliding window) are often equally great, but sometimes fall off a cliff. See my plot of Rsqr (correlation of future ten year return and present CAPE) as you slide a 15 year window back in time.

I'll have more on increasing multiples and other things in Part II.